CN109581194B - A Dynamic Generation Method of Electronic System Fault Testing Strategy - Google Patents

Abstract

The invention discloses a method for dynamically generating a fault test strategy of an electronic system. According to the statically generated search tree and the change amount of dynamic test points or faults, the cost of the node search strategy in the search tree is dynamically modified, and each fault in the search strategy is updated. The optimal test point of the fuzzy set is then quickly updated to obtain an optimized test plan, which improves the optimization efficiency of the overall test and reduces the number of searches.

Description

A Dynamic Generation Method of Electronic System Fault Testing Strategy

technical field

The invention belongs to the technical field of circuit fault diagnosis, and more particularly, relates to a dynamic generation method of an electronic system fault test strategy.

Background technique

With the development of electronic technology, the testability design of electronic system has become an important part in the field of electronic design. Among them, timely and accurate determination of circuit status and isolation of internal faults can effectively shorten the development, experiment and release time of electronic systems. Therefore, the design of fault test scheme is of great significance in practical applications.

In the existing fault test scheme design method, the sequential test is based on the signal flow diagram given by the preliminary design and the circuit relationship described by the correlation model, and the test sequence test method is given, which reduces the cost of the test and can effectively improve the later stage. Efficiency of design and verification evaluations, therefore, this technique is widely used in the design of electronic systems for testability.

However, in the actual test process, the test results may lead to the change of the fault propagation model and affect the test process. Therefore, the dynamic test optimization method has extremely high practical value in the research of sequential test.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the deficiencies of the prior art, and to provide a method for dynamically generating a fault test strategy for an electronic system, which generates a fault diagnosis method in a dynamic state based on the fault dependency information of the electronic system and a static search strategy, and has the advantages of low optimization strategy cost, Features such as fast search speed.

In order to achieve the above-mentioned purpose of the invention, the present invention proposes and implements a method for dynamically generating an electronic system fault test strategy, which is characterized in that it includes the following steps:

(1), build a test model

According to the relationship between the internal fault state of the circuit and the output of the measuring points in the circuit, the test model H={D,C,P} of the circuit is constructed, where C={c ₁ ,c ₂ ,...,ci ,...c _{M }} is The matrix formed by the cost corresponding to the test points set in the circuit, c _i represents the a priori test cost of the i-th test point in the circuit, M is the total number of optional test points; P={p ₁ ,p ₂ ,...,p _j ,...,p _N } is the matrix formed by the probability of occurrence of each fault state in the circuit system, p _j is the probability of occurrence of the jth fault state, N is the total number of fault states; D is the fault dependency matrix, specifically expressed as:

Among them, d _ij represents the test information of the j-th circuit fault at the i-th measuring point, the value of d _ij is 0 or 1, and d _ij =1 indicates that the j-th circuit fault can be detected by the i-th measuring point , d _ij =0 means that the jth fault cannot be detected by the ith measuring point;

(2), build the initial search tree

An initial search tree T={T ₁ ,T ₂ ,…,T _p ,…,T _P } is constructed by a static heuristic search method, where T _p represents the p-th node in the search tree, and T ₁ is the search tree. The root node of the tree, P is the total number of nodes in the search tree;

For a node T _p in the search tree, there is T _p ={S _r ,t _opt ,Cost _ava ,S ₀ ,S ₁ }, where S _r is the faulty fuzzy set of the node, S _r ={s _r1 ,s _r2 ,…,s _rj ,…,s _rM }, s _rj is the fault state contained in the fault fuzzy set; to _opt is the optimal measurement point in the search tree; t _ava ={t _ava1 ,t _ava2 ,…,t _avai , ...,t _avaL } is the set of valid measurement points corresponding to S _r , t _avai is the measurement point that effectively separates S _r , and L is the number of valid measurement points; Cost _ava = {Cost _ava1 ,Cost _ava2 ,...,Cost _avai ,...,Cost _avaL } is the cost of the test method generated by the corresponding valid measurement point, Cost _avai is the cost of the test method generated by the corresponding valid measurement point t _avai ; S ₀ and S ₁ are the sub-fault fuzzy obtained by dividing S _r according to to _opt set, which is defined as:

及该测点对应的代价变化量

或该测得对应的更新故障状态

及故障概率

(3), read in the updated measuring point

and the cost change corresponding to the measurement point

or the update fault state corresponding to this measurement

and failure probability

(4), take the root node T ₁ of the search tree as the initial node of the update process, and take the fault fuzzy set S _r of the root node as the target fuzzy set S of the node to be updated in the search tree;

(5), update the fault tree according to the fault dependency matrix

若t_ava中包含更新的测点

则进行步骤(5.2)，否则跳至步骤(6)；(5.1), determine whether the valid measurement point set _tava of the node to be updated contains updated measurement points

If t _ava contains updated measuring points

Then go to step (5.2), otherwise skip to step (6);

若t_optt为更新的测点

则进入步骤(5.3)；否则跳至步骤(5.6)；(5.2), determine whether the current optimal measurement point t _opt of the node to be updated is the updated measurement point

If t _optt is the updated measuring point

Then go to step (5.3); otherwise, skip to step (5.6);

(5.3), the total cost of updating the fault strategy of the optimal measurement point to _opt

Among them, _{pi is the probability corresponding to the fault state si ;}

(5.4), update the fault strategy cost generated by the other valid measurement points except the optimal measurement point to _opt ;

(5.4.1), set the effective measuring point t' as the first effective measuring point in _tava ;

(5.4.2), if t' is the current optimal measurement point _tot , go to step (5.4.7); otherwise, according to the fault dependency information of t', separate the fault set S into the left sub fault set S _{t' ,0} and the right sub-fault set S _t',1 :

S _t',0 ={s _j |d _jt' =0}

S _t',1 ={s _j |d _jt' =1}

(5.4.3), find the left child node corresponding to the fault set S _t',0 in the search tree, take this node as the node to be updated, and then calculate the left child according to the methods described in steps (5.1) to (5.3). The optimal solution cost corresponding to the node is Costt', 0;

(5.4.4), find the right child node corresponding to the fault set S _t',1 in the search tree, take this node as the node to be updated, and then calculate the right child according to the methods described in steps (5.1) to (5.3). The optimal solution cost Costt', 1 corresponding to the node;

(5.4.5), update t' with regard to the optimal solution cost generated by the fault set S:

(5.4.6), determine whether the corresponding costs of all measuring points in t _ava are updated, if all are updated, go to step (5.5); otherwise, set t' as the next valid measuring point to be updated in t _ava , and then Return to step (5.4.2);

(5.5), record the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point _tot , and set its corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set of the node in the search tree Set S ₀ and right sub-fuzzy set S ₁ , then enter step (6);

对应的故障诊断总代价:(5.6), update the measuring points of the cost change

The corresponding total cost of fault diagnosis:

为

的初始代价，

和

分别为左子故障集和右子故障集的测试代价；in,

for

the initial cost of

and

are the test costs of the left sub-fault set and the right sub-fault set, respectively;

(5.7), update the fault strategy cost generated by the remaining valid measurement points except the cost change measurement point;

(5.7.1), set the effective measuring point t' as the first effective measuring point in _tava ;

则进行步骤(5.7.6)；否则，根据t'的故障依赖信息，将故障集S分离为左子故障集S_t',0和右子故障集S_t',1:(5.7.2), if t' is

Then go to step (5.7.6); otherwise, according to the fault dependency information of t', the fault set S is separated into the left sub-fault set S _t',0 and the right sub-fault set S _t',1 :

S _t',0 ={s _j |d _jt' =0}

S _t',1 ={s _j |d _jt' =1}

(5.7.3), find the left child node corresponding to the fault set S _t',0 in the search tree, take this node as the node to be updated, and then calculate the left child according to the methods described in steps (5.1) to (5.3). The optimal solution cost corresponding to the node is Costt', 0;

(5.7.4), find the right child node corresponding to the fault set S _t',1 in the search tree, take this node as the node to be updated, and then calculate the right child according to the methods described in steps (5.1) to (5.3). The optimal solution cost Costt', 1 corresponding to the node;

(5.7.5), update the optimal solution cost generated by t' with respect to the fault set S:

(5.7.6), determine whether the corresponding costs of all measuring points in t _ava are updated, if all are updated, go to step (5.8); set t' as the next valid measuring point to be updated in t _ava , and then return step (5.7.2);

(5.8), take the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point _tot , and set its corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set of the node in the search tree S ₀ and the right sub-fuzzy set S ₁ , then enter step (6);

若更新节点的故障集中包含待更新的故障状态

则进行步骤(7)，否则跳转至步骤(9)；(6), determine whether the fault set of the node to be updated contains the updated fault state

If the fault set of the update node contains the fault status to be updated

Then go to step (7), otherwise jump to step (9);

(7), update the test cost corresponding to the available measurement point _tava

(7.1), set the effective measuring point t' as the first effective measuring point in _tava ;

(7.2) According to the fault dependency information of t', separate the fault set S into the left sub-fault set S _t',0 and the right sub-fault set S _t',1

S _t',0 ={s _j |d _jt' =0}

S _t',1 ={s _j |d _jt' =1}

则将该节点作为待更新节点，然后按照步骤(5.1)～(5.3)所述方法计算出左子节点对应的最优解代价Costt',0；否则将搜索树中的左子节点的代价作为最优解代价Costt',0；(7.3), find the left child node corresponding to the fault set S _t',0 in the search tree, if S _t',0 contains the updated fault state

Then take the node as the node to be updated, and then calculate the optimal solution cost Costt', 0 corresponding to the left child node according to the methods described in steps (5.1) to (5.3); otherwise, take the cost of the left child node in the search tree as The optimal solution cost Costt', 0;

则将该节点作为待更新节点，然后按照步骤(5.1)～(5.3)所述方法计算出右子节点对应的最优解代价Costt',1；否则将搜索树中的左子节点的代价作为最优解代价Costt',1；(7.4) Find the right child node corresponding to the fault set S _t',1 in the search tree, if S _t',1 contains the updated fault state

Then take the node as the node to be updated, and then calculate the optimal solution cost Costt',1 corresponding to the right child node according to the methods described in steps (5.1) to (5.3); otherwise, take the cost of the left child node in the search tree as The optimal solution cost Costt', 1;

(7.5), update t' to generate the optimal solution cost of fault set S:

(7.6), if the corresponding cost of all valid measurement points in t _ava is updated, then enter step (8), otherwise set t' to the next valid measurement point to be updated in t _ava , and then return to step (7.2);

(8) Update the optimal node information and optimal cost of the current effective node. After the update is completed, set the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point _tot , and set the corresponding left child The fuzzy set and the right sub-fuzzy set are set as the left sub-fuzzy set S ₀ and the right sub-fuzzy set of the optimal measurement point in the search tree;

(9), return the optimal node information and optimal test cost of the current optimal measurement point, so as to dynamically generate the electronic system fault test strategy.

The purpose of the invention of the present invention is achieved in this way:

The present invention is a method for dynamically generating a fault testing strategy of an electronic system. According to the statically generated search tree and the changes of dynamic test points or faults, the cost of the node search strategy in the search tree is dynamically modified, and each fault fuzzy set in the search strategy is updated. The optimal test point is then quickly updated to obtain an optimized test plan, which improves the optimization efficiency of the overall test and reduces the number of searches.

At the same time, a method for dynamically generating an electronic system fault test strategy of the present invention also has the following beneficial effects:

(1) The present invention corrects the static diagnosis method according to the change in the fault model in the dynamic state. Compared with the traditional method of rebuilding the diagnosis tree, it can better utilize the prior knowledge obtained in the static search, and effectively improve the generation of optimal Efficiency of diagnostic trees.

(2) In the present invention, by screening nodes in the search process, the search of nodes not affected by the change amount is avoided, the search space is reduced, and the search efficiency is improved.

Description of drawings

Fig. 1 is a kind of flow chart of the dynamic generation method of electronic system fault test strategy of the present invention;

Fig. 2 is the diagnosis tree generated by the present invention after measuring point cost is changed;

Fig. 3 is a diagnosis tree generated by the present invention after the failure probability is changed.

Detailed ways

The specific embodiments of the present invention are described below with reference to the accompanying drawings, so that those skilled in the art can better understand the present invention. It should be noted that, in the following description, when the detailed description of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here.

Example

FIG. 1 is a flow chart of a method for dynamically generating a fault test strategy for an electronic system according to the present invention.

In this embodiment, as shown in FIG. 1 , a method for dynamically generating a fault test strategy for an electronic system of the present invention includes the following steps:

S1. Build a test model

According to the relationship between the internal fault state of the circuit and the output of the measuring points in the circuit, the test model H={D,C,P} of the circuit is constructed, where C={c ₁ ,c ₂ ,...,ci ,...c _{M }} is The matrix formed by the cost corresponding to the test points set in the circuit, c _i represents the a priori test cost of the i-th test point in the circuit, M is the total number of optional test points; P={p ₁ ,p ₂ ,...,p _j ,...,p _N } is the matrix formed by the probability of occurrence of each fault state in the circuit system, p _j is the probability of occurrence of the jth fault state, N is the total number of fault states; D is the fault dependency matrix, specifically expressed as:

Among them, d _ij represents the test information of the j-th circuit fault at the i-th measuring point, the value of d _ij is 0 or 1, and d _ij =1 indicates that the j-th circuit fault can be detected by the i-th measuring point , d _ij =0 means that the jth fault cannot be detected by the ith measuring point;

S2. Build the initial search tree

An initial search tree T={T ₁ ,T ₂ ,…,T _p ,…,T _P } is constructed by a static heuristic search method, where T _p represents the p-th node in the search tree, and T ₁ is the search tree. The root node of the tree, P is the total number of nodes in the search tree;

For a node T _p in the search tree, there is T _p ={S _r ,t _opt ,Cost _ava ,S ₀ ,S ₁ }, where S _r is the faulty fuzzy set of the node, S _r ={s _r1 ,s _r2 ,...,s _rj ,...,s _rM }, s _rj is the j-th fault state included in the fault fuzzy set; to _opt is the optimal measurement point in the search tree; t _ava ={t _ava1 ,t _ava2 ,..., t _avai ,...,t _avaL } is the set of valid measurement points corresponding to S _r , t _avai is the measurement point that effectively separates S _r , and L is the number of valid measurement points; Cost _ava = {Cost _ava1 ,Cost _ava2 ,… ,Cost _avai ,...,Cost _avaL } is the cost of the test method generated by the corresponding valid measurement point, Cost _avai is the cost of the test method generated by the corresponding valid measurement point t _avai ; S ₀ and S ₁ are obtained by dividing S _r according to to _opt Subfault fuzzy set, which is defined as:

及该测点对应的代价变化量

或该测得对应的更新故障状态

及故障概率

S3. Read in the updated measuring point

and the cost change corresponding to the measurement point

or the update fault state corresponding to this measurement

and failure probability

S4, take the root node T1 _of the search tree as the initial node of the update process, and take the fault fuzzy set S _r of the root node as the target fuzzy set S of the node to be updated in the search tree;

S5. Update the fault tree according to the fault dependency matrix

若t_ava中包含更新的测点

则进行步骤S5.2，否则跳至步骤S6；S5.1. Determine whether the valid measurement point set _tava of the node to be updated contains updated measurement points

If t _ava contains updated measuring points

Then go to step S5.2, otherwise skip to step S6;

若t_optt为更新的测点

则进入步骤S5.3；否则跳至步骤S5.6；S5.2. Determine whether the current optimal measurement point t _opt of the node to be updated is the updated measurement point

If t _optt is the updated measuring point

Then go to step S5.3; otherwise, skip to step S5.6;

S5.3. The total cost of updating the fault strategy of the optimal measurement point to _opt

Among them, _{pi is the probability corresponding to the fault state si ;}

S5.4, update the cost of the fault strategy generated by the other valid measurement points except the optimal measurement point to _opt ;

_S5.4.1 . Set the effective measuring point t' as the first effective measuring point in tava;

S5.4.2. If t' is the current optimal measurement point tot , go to step _S5.4.7 ; otherwise, according to the fault dependency information of t', separate the fault set S into left sub-fault sets S _t',0 and right Sub-fault set S _t',1 :

S _t',0 ={s _j |d _jt' =0}

S _t',1 ={s _j |d _jt '=1}

S5.4.3. Find the left child node corresponding to the fault set S _t ', ₀ in the search tree, take the node as the node to be updated, and then calculate the corresponding left child node according to the method described in steps S5.1 to S5.3 The optimal solution cost Costt',0;

S5.4.4. Find the right child node corresponding to the fault set S _t ', ₁ in the search tree, take the node as the node to be updated, and then calculate the corresponding right child node according to the method described in steps S5.1 to S5.3 The optimal solution cost Costt',1;

S5.4.5. Update t' to generate the optimal solution cost of fault set S:

S5.4.6. Determine whether the corresponding costs of all measuring points in t _ava have been updated, if all, go to step S5.5; otherwise, set t' as the next valid measuring point to be updated in t _ava , and then return to step S5.4.2;

_S5.5 . Record the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point tot , and set its corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set of the node in the search tree Set S ₀ and right sub-fuzzy set S ₁ , then enter step S6;

对应的故障诊断总代价:S5.6, update the measuring points of the cost change

The corresponding total cost of fault diagnosis:

为

的初始代价，

和

分别为左子故障集和右子故障集的测试代价；in,

for

the initial cost of

and

are the test costs of the left sub-fault set and the right sub-fault set, respectively;

S5.7. Update the fault strategy cost generated by the remaining valid measurement points except the cost change measurement point;

_S5.7.1 . Set the effective measuring point t' as the first effective measuring point in tava;

则进行步骤S5.7.6；否则，根据t'的故障依赖信息，将故障集S分离为左子故障集S_t',₀和右子故障集S_t',₁:S5.7.2, if t' is

Then go to step S5.7.6; otherwise, according to the fault dependency information of t', the fault set S is separated into the left sub-fault set S _t ', ₀ and the right sub-fault set S _t ', ₁ :

S _t ', ₀ ={s _j |d _jt '=0}

S _t ', ₁ ={s _j |d _jt '=1}

S5.7.3. Find the left child node corresponding to the fault set S _t ', ₀ in the search tree, take the node as the node to be updated, and then calculate the corresponding left child node according to the method described in steps S5.1 to S5.3 The optimal solution cost Costt',0;

S5.7.4. Find the right child node corresponding to the fault set S _t',1 in the search tree, take this node as the node to be updated, and then calculate the corresponding right child node according to the methods described in steps S5.1 to S5.3 The optimal solution cost Costt',1;

S5.7.5. Update t' to generate the optimal solution cost of fault set S:

S5.7.6, determine whether the corresponding costs of all measuring points in t _ava are updated, if all are updated, then go to step S5.8; set t' as the next valid measuring point to be updated in t _ava , and then return to step S5 .7.2;

_S5.8 . Set the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point tot , and set its corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set of the node in the search tree S ₀ and the right sub-fuzzy set S ₁ , then enter step S6;

若更新节点的故障集中包含待更新的故障状态

则进行步骤S7，否则跳转至步骤S9；S6. Determine whether the fault set of the node to be updated contains the updated fault state

If the fault set of the update node contains the fault status to be updated

Then go to step S7, otherwise jump to step S9;

S7, update the test cost corresponding to the available test point _tava

_S7.1 . Set the effective measuring point t' as the first effective measuring point in tava;

S7.2. According to the fault dependency information of t', separate the fault set S into a left sub-fault set S _t',0 and a right sub-fault set S _t',1

S _t',0 ={s _j |d _jt' =0}

S _t',1 ={s _j |d _jt' =1}

则将该节点作为待更新节点，然后按照步骤S5.1～S5.3所述方法计算出左子节点对应的最优解代价Costt',0；否则将搜索树中的左子节点的代价作为最优解代价Costt',0；S7.3. Find the left child node corresponding to the fault set S _t',0 in the search tree, if S _t',0 contains the updated fault state

Then take the node as the node to be updated, and then calculate the optimal solution cost Costt',0 corresponding to the left child node according to the method described in steps S5.1 to S5.3; otherwise, take the cost of the left child node in the search tree as The optimal solution cost Costt', 0;

则将该节点作为待更新节点，然后按照步骤S5.1～S5.3所述方法计算出右子节点对应的最优解代价Costt',1；否则将搜索树中的左子节点的代价作为最优解代价Costt',1；S7.4. Find the right child node corresponding to the fault set S _t',1 in the search tree, if S _t',1 contains the updated fault state

Then take the node as the node to be updated, and then calculate the optimal solution cost Costt',1 corresponding to the right child node according to the method described in steps S5.1 to S5.3; otherwise, take the cost of the left child node in the search tree as The optimal solution cost Costt', 1;

S7.5. Update t' to generate the optimal solution cost of the fault set S:

S7.6, if the corresponding costs of all valid measuring points in t _ava are updated, then enter step S8, otherwise set t' as the next valid measuring point to be updated in t _ava , and then return to step S7.2;

S8. Update the optimal node information and optimal cost of the current effective node. After the update is completed, set the measurement point corresponding to the minimum value in Cost _ava as the new optimal measurement point _tot , and set its corresponding left sub-fuzzy set and the right sub-fuzzy set are set as the left sub-fuzzy set S ₀ and the right sub-fuzzy set of the optimal measurement point in the search tree;

S9. Return the optimal node information and optimal test cost of the current optimal measurement point, so as to dynamically generate an electronic system fault test strategy.

example

In order to illustrate the technical effect of the present invention, an anti-tank system is used as an example to verify the present invention. Its mechatronic part involves a total of 13 system states and 12 available measuring points, a fault dependency matrix, a priori probability corresponding to each system state, and The test cost of each measuring point is shown in Table 1. The cost value of t ₁ , t ₇ and the occurrence probability of s ₃ and s ₆ are respectively changed to simulate the change of parameters in the dynamic process. In order to verify the effect of the algorithm proposed by the present invention, the anti-tank system is selected as an example. The instance is computed together as a contrast algorithm.

Table 1 is the failure-dependency matrix of the anti-tank system;

Table 1

When c1 is increased by 0.8, the diagnosis tree constructed by the present invention is shown in Fig. 2. It can be seen from Fig. 2 that the present invention generates a diagnosis method that can effectively isolate all faults. Table 2 shows the total cost of the fault tree generated by AO* and this method when the cost of the measurement point changes. It can be obtained from Table 2 that both AO* and the present invention can generate the optimal diagnostic method of the test cost under the condition of the change of the test point cost. However, the present invention has obvious advantages over the AO* algorithm in generation time.

Table 2

When p3 is reduced by 0.02, the diagnostic tree constructed by the present invention is shown in FIG. 3 . It can be seen from FIG. 3 that the present invention generates a diagnostic method that can effectively isolate all faults. Table 3 shows the total cost and generation time of the fault tree generated by AO* and this method when the fault probability changes. It can be obtained from Table 3 that both AO* and the present invention can generate a diagnostic method with the optimal test cost under the condition that the test point cost changes. However, the present invention has obvious advantages over the AO* algorithm in generation time.

table 3

Although the illustrative specific embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, As long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the inventive concept are included in the protection list.

Claims (1)

1. A dynamic generation method for an electronic system fault test strategy is characterized by comprising the following steps:

(1) and constructing a test model

According to the relation between the internal fault state of the circuit and the output of a measuring point in the circuit, a test model H of the circuit is constructed, wherein C is { D, C, P }, and C is { C }₁,c₂,…,c_i,…c_MIs a generation corresponding to a measuring point arranged in the circuitA matrix of valence formations, c_iThe prior test cost of the ith test point in the circuit is represented, and M is the total number of the selectable test points; p ═ P₁,p₂,…,p_j,…,p_NIs a matrix formed by the probability of each fault state in the circuit system, p_jThe probability of occurrence of the jth fault state is shown, and N is the total number of the fault states; d is a fault dependency matrix, which is specifically expressed as:

wherein d is_ijRepresents the test information of the jth circuit fault at the ith test point, d_ijIs 0 or 1, d_ij1 means that the jth circuit fault can be detected at the ith test point, d_ij0 represents that the jth fault cannot be detected by the ith measuring point;

(2) building an initial search tree

Constructing an initial search tree T ═ { T ] through a static heuristic search method₁,T₂,…,T_p,…,T_PIn which T is_pRepresenting the p-th node, T, in the search tree₁Is the root node of the search tree, and P is the total number of nodes in the search tree;

for node T in search tree_pHaving a value of T_p＝{S_r,t_opt,Cost_ava,S₀,S₁In which S is_rIs a fuzzy set of faults of a node, S_r＝{s_r1,s_r2,…,s_rj,…,s_rM}，s_rjFault states contained in the fuzzy set of faults; t is t_optSearching for an optimal measuring point in the tree; t is t_ava＝{t_ava1,t_ava2,…,t_avai,…,t_avaLIs S_rSet of corresponding valid measurement points, t_avaiTo separate S effectively_rL is the number of effective measuring points; cost_ava＝{Cost_ava1,Cost_ava2,…,Cost_avai,…,Cost_avaLCost of the test method generated for the corresponding valid test points, Cost_avaiTo correspond to the effective measuring point t_avaiCost of the generated test method; s₀And S₁Is according to t_optSplitting S_rThe resulting fuzzy set of sub-faults, which is defined as:

(3) reading in the updated measurement point

And the cost variation corresponding to the measuring point

Or the updated fault state corresponding to the measuring point

And probability of failure

(4) Will search the root node T of the tree₁As an initial node of the updating process, a fault fuzzy set S of a root node is set_rA target fuzzy set S as a node to be updated in the search tree;

(5) updating the fault tree according to the fault dependency matrix

(5.1) judging the effective measuring point set t of the node to be updated_avaWhether or not to include updated measurement points

If t_avaIncluding updated measurement points

Then step (5.2) is carried out, otherwise step is jumped to(6)；

(5.2) judging the current optimal measuring point t of the node to be updated_optWhether it is an updated measurement point

If t_opttFor updated measuring point

Entering step (5.3); otherwise, jumping to the step (5.6);

(5.3) updating the optimal measuring point t_optTotal cost of fault strategy of

Wherein p is_iIs a fault state s_iA corresponding probability;

(5.4) updating and dividing the optimal measuring point t_optThe cost of the fault strategy generated by the other effective measuring points;

(5.4.1) the effective measuring point t' is t_avaThe first active station in (a);

(5.4.2) if t' is the current optimal measuring point t_optThen go to step (5.4.7); otherwise, according to the fault dependency information of t', the fault set S is separated into a left sub-fault set S_t',0And right sub-fault set S_t',1:

S_t',0＝{s_j|d_jt'＝0}

S_t',1＝{s_j|d_jt'＝1}

(5.4.3) finding the fault set S in the search tree_t',0Taking the corresponding left child node as a node to be updated, and then calculating the optimal solution cost Costt' 0 corresponding to the left child node according to the method in the steps (5.1) - (5.3);

(5.4.4) finding the fault set S in the search tree_t',1Corresponding right child nodeTaking the node as a node to be updated, and then calculating the optimal solution cost Costt' 1 corresponding to the right child node according to the methods in the steps (5.1) - (5.3);

(5.4.5), updating t' the optimal solution cost generated by the fault set S:

(5.4.6) determination of t_avaWhether the corresponding costs of all the measuring points are updated or not is judged, and if the corresponding costs of all the measuring points are updated, the step (5.5) is carried out; otherwise, set t' to t_avaReturning to the step (5.4.2) when the next effective measuring point to be updated is reached;

(5.5) Cost_avaThe measuring point corresponding to the minimum value is marked as a new optimal measuring point t_optAnd setting the corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set S of the node in the search tree₀And the right sub-fuzzy set S₁Then entering the step (6);

(5.6) updating the measurement points with changed cost

The corresponding total cost of fault diagnosis:

wherein,

is composed of

The initial cost of (a) of (b),

and

are each left childTesting costs of the fault set and the right sub-fault set;

(5.7) updating the fault strategy cost generated by the other effective measuring points except the cost change measuring point;

(5.7.1) the effective measuring point t' is t_avaThe first active station in (a);

(5.7.2) if t' is

Then step (5.7.6) is performed; otherwise, according to the fault dependency information of t', the fault set S is separated into a left sub-fault set S_t',0And right sub-fault set S_t',1:

S_t',0＝{s_j|d_jt'＝0}

S_t',1＝{s_j|d_jt'＝1}

(5.7.3) finding the fault set S in the search tree_t',0Taking the corresponding left child node as a node to be updated, and then calculating the optimal solution cost Costt' 0 corresponding to the left child node according to the method in the steps (5.1) - (5.3);

(5.7.4) finding the fault set in the search tree as S_t',1Taking the corresponding right child node as a node to be updated, and then calculating the optimal solution cost Costt' 1 corresponding to the right child node according to the method in the steps (5.1) - (5.3);

(5.7.5), updating t' the optimal solution cost generated about the fault set S:

(5.7.6) determination t_avaWhether the corresponding costs of all the measuring points are updated or not is judged, and if the corresponding costs of all the measuring points are updated, the step (5.8) is carried out; setting t' to t_avaReturning to the step (5.7.2) when the next effective measuring point to be updated is reached;

(5.8) Cost_avaThe measuring point corresponding to the minimum value is a new optimal measuring point t_optAnd setting the corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set of the node in the search treeCollection S₀And the right sub-fuzzy set S₁Then entering the step (6);

(6) judging whether the fault set of the nodes to be updated contains the updated fault state

If the fault set of the updating node contains the fault state to be updated

Step (7) is carried out, otherwise, step (9) is skipped;

(7) updating the available measuring point t_avaCorresponding test cost

(7.1) setting an effective measuring point t' as t_avaThe first active station in (a);

(7.2) according to the fault dependency information of t', separating the fault set S into a left sub-fault set S_t',0And right sub-fault set S_t',1

S_t',0＝{s_j|d_jt'＝0}

S_t',1＝{s_j|d_jt'＝1}

(7.3) finding the fault set S in the search tree_t',0Corresponding left child node, if S_t',0Including updated fault conditions

Taking the node as a node to be updated, and then calculating the optimal solution cost Costt' 0 corresponding to the left child node according to the method in the steps (5.1) - (5.3); otherwise, taking the cost of the left child node in the search tree as the optimal solution cost Costt', 0;

(7.4) finding the fault set S in the search tree_t',1The corresponding right child node, if S_t',1Including updated fault conditions

Then the node is taken as the node to be updated, and then the right child node is calculated according to the methods of the steps (5.1) - (5.3)The corresponding optimal solution cost Costt', 1; otherwise, taking the cost of the left child node in the search tree as the optimal solution cost Costt', 1;

(7.5) updating t' about the optimal solution cost generated by the fault set S:

(7.6) if t_avaIf all the cost corresponding to the effective measuring points are updated, the step (8) is carried out, otherwise, t' is set as t_avaThen returning to the step (7.2);

(8) updating the optimal node information and the optimal Cost of the current effective node, and after the updating is finished, using the Cost_avaSetting the measuring point corresponding to the minimum value as a new optimal measuring point t_optAnd setting the corresponding left sub-fuzzy set and right sub-fuzzy set as the left sub-fuzzy set S of the optimal measuring point in the search tree₀And a right sub-fuzzy set;

(9) and returning the optimal node information and the optimal test cost of the current optimal test point, thereby dynamically generating a fault test strategy of the electronic system.

Images